Welcome to iraf.net Thursday, March 28 2024 @ 08:20 PM GMT
juge |
04/13/2017 09:41AM (Read 1386 times)
|
|
|
Status: offline
Registered: 04/13/2017
Posts: 1
|
Hi guys,
I use the psf task in daophot package to build semi-emperical PSFs (Gaussians+residuals) for my images. One thing that puzzles me is that when the gain parameter is arbitrarily adjusted, output of the psf task does not change at all. If the task fits gaussian function to flux distribution weighted by poisson noise plus some other noise (e.g. readout noise), then the gain parameter should have some impacts on the best-fit parameters of the gaussian function. I would like to know the detailed weighting scheme during the fitting process. Any documentation or discussion to which I can refer?
Thanks in advance!
|
|
|
|
valdes |
04/25/2017 10:04PM
|
|
|
Status: offline
Registered: 11/11/2005
Posts: 728
|
Hello,
The author of the IRAF version of the daophot package and the psf task is no longer available. I am the only IRAF science programmer with the core IRAF group. I have not had occasion to use this task.
However, a quick probe of the help page and the source code leads me to conclude that the gain parameter is not used by the psf task. This task, as well as others in the daophot package, include a link to the "datapars" parameter set so that this can be set once and apply to various tasks on the same type of data. In this case the gain value does not seem to come into play for psf. The help page just indicates that the parameters used from the datapars parameter set are scale, datamin, and datamax. Now you are probably right that the noise properties of the data should come into play for a complete best fit. Since I don't much about the algorithms or goals of this task I can't offer much more of a response.
Frank Valdes
|
|
|
|
| |
|
Content generated in: 0.08 seconds |
|